Go Beyond Traditional Data To Feed Next-Gen AI

• Read original article here

When AI was introduced to the business community, visions of how we could use it to meet customers’ needs came clearly into focus. And, with those visions, came the promise for AI to revolutionize customer experience. The acquisition and successful management of data were early stumbling blocks on our collective journey toward that future. But now, organizations are getting better at managing data and leveraging it too. Still, the visions of perfecting the customer experience with AI remain barely out of reach.

To achieve this higher-order AI, the progressive companies are asking, “What’s next?” In addition to the typical approach of sucking up more data and managing it better, one needs to consider more innovative approaches. While data maturity varies wildly across companies and industries, I’m starting to see transformational uses of data at the cutting edge. Companies can use these innovative data assets (or create their own) to power more advanced AI. Innovations like these introduce data where none existed before, use insights as data and reach for data beyond the typical customer journey.

Companies are finding ways to create data where none previously existed and making it possible to meet needs no one previously considered. The practice of finding or creating data to solve a known problem is not new. But what if you start with a novel data source and then think of novel ways you can use that data?

For example, a company called Label Insight captures attributes for food and beverage products directly from the labels. This goes beyond logistics data, like product size and weight, and includes ingredients, nutrition and marketing claims. Now Label Insight offers a centralized source that dimensionalizes the entire grocery store for anyone who wants to subscribe. What could a company do with this data? Use it to create a recommendation service that suggests products to consumers the way Netflix suggests shows to watch? What about a service that builds weekly menus and shopping lists based on a user’s nutritional requirements?

In healthcare, Trialtrove is a source for pharmaceutical clinical trial data. It uses several thousand sources to deliver the most comprehensive trial intelligence. The possibilities are endless, and they’re enabled by a simple impulse to make new data and see where it leads.

We think of data, then we think of insights from data. But what happens when we think of insights as the data? When you’ve gathered enough insights on something, it’s possible to use those to drive more impressive conclusions. Knowledge graphs are an example of explicit representation of the insights and knowledge that exist in the data. But when the databases are large and have many dimensions, it may not be feasible to have such explicit representations and more sophisticated methods are needed.

Customer insights are a rich example. By using an extensive source of customer insights as derivative data, the possibility for next-generation customer understanding becomes easier and, in some cases, more routinely possible. As an illustration, consider starting with a 360-degree view of your customers. This data likely includes personal attributes, customer actions, social and other network data, demographics, personal preferences and more. From this, we can generate insights such as propensity to interact with a channel and propensity to adopt a product. When these insights are themselves used as data, we not only understand our customers better, but we’re also able to predict their behavior more accurately. As we add more information, such representations increase in accuracy. With this solution, every targeted investment you would make to reach the right customers would become more and more reliable as the data accumulated.

Companies have been studying and mapping their customer journeys for some time now. We know customer experience is king, so it’s essential to know everything possible about a customer’s purchase journey from beginning to end. But what if you thought beyond your own company’s data and incorporated interrelated activities and purchases from other companies to improve customer experience?

Journera is an example of a platform for the travel industry that helps companies leverage traveler data. Airlines, for example, can factor in a flyer’s car rentals and hotel stays. Wouldn’t a customer’s experience be better if a flight delay led to quick and easy modifications to their car rental and hotel bookings?

This kind of solution could have many applications. What about getting help with all the many decisions and purchases that go into a home remodel? Could back-to-school shopping be simplified?

Of course, we should continue to acquire more data, better data and make continual improvements to its stewardship. But a new generation of possibilities exists if we’re able to think outside the old data acquisition and management paradigm. As more of the world becomes digitized, insights from insights become prevalent and the siloes we place around customer journeys will continue to disappear. That’s the future we first envisioned when we learned what AI could do. And that future is already here. We just need a little extra imagination to access it.